What do /a/ and /O/ merge to?

[Beverly Flanigan]

Southern Ohio is like southern Indiana, and I've written on this in an LVC
article a couple years ago (look under Flanigan and Norris 2000).  This
merger almost to /O/ (but not quite--it's actually between /a/ and /O/, the
"turned script a" of Kenyon, Kurath and McDavid, and others) starts in
western PA and spreads southwestward along the Ohio River; it's in
Youngstown and downward.  But it's also in the Toronto area (essentially
the same, though a bit more "British").  So it's a merger, as Labov says in
the Atlas, but he doesn't deal with the "merger toward what" issue, as I
do.  A good test I've used with my regional students is to ask them to
transcribe 'body', 'doll', 'on' and 'polish' as well as the usual 'cot' and
'caught'--and they typically use /O/ for all of them.  (I teach them basic
IPA, without all the in-between niceties, so they have a forced choice
between /a/ and /O/.)  A couple of people have suggested that western PA
has lost the phoneme /O/ but has kept a phonemic distinction between /a/
and /turned script a/, and I agree for this area; /a/ is of course used in
'father' and other words of that set.

Beverly Flanigan
Ohio University

At 03:07 PM 3/4/2003 -0500, you wrote:
>When I talk about the /a/~/O/ merger in various of my classes I fairly
>regularly find that among students who merge them, some merge to /a/ and
>others to /O/.  Based just on observing my students, it appears that
>students from southern Indiana, south of Indianapolis, who merge are like to
>merge to /O/.  Students from north of Indy merge to /a/.  But I haven't
>collected specific data on this.  I've looked through the discussion on the
>Phonological Atlas of North America site, although I haven't read closely
>the newly revised Chapter 11 that has to be the most extensive treatment of
>the merger anywhere.  However, I haven't come across the question of what
>they merge to.  Has anyone investigated this?
>
>Herb